Objective: To measure umbilical cord blood oxygen saturation, to calculate preductal oxygen saturation at birth, and to assess its predictive value for acidosis.
Methods: Umbilical cord blood samples of 1537 live-born singleton neonates were analyzed. Oxygen saturation was measured by spectrophotometry; pH and base excess were measured by a pH and blood gas analyzer. Preductal oxygen saturation was calculated with an empirical equation. Acidosis was defined as 2 standard deviations (SDs) below the mean of umbilical artery (UA) pH or base excess (7.09 and -10.50 mmol/L, respectively). The predictive value for acidosis of UA and umbilical vein (UV) oxygen saturation and calculated preductal oxygen saturation was determined with receiver operating characteristic curves.
Results: The mean values (+/-SD) of UV, UA, and calculated preductal oxygen saturation were 52 +/- 18%, 26 +/- 17%, and 31 +/- 16%, respectively. Forty-seven neonates had UA pH less than 7.09 and 60 had UA base excess less than -10.50 mmol/L. The UV, UA, and calculated preductal oxygen saturation showed considerably weaker relations to UA base excess (multiple r(2) =.056,.003, and.017, respectively; P <.001) than to UA pH (multiple r(2) =.112,.126, and.148, respectively; P <. 001). Receiver operating characteristic areas under the curve were higher when predicting low pH compared with low base excess (for UV, UA, and calculated preductal oxygen saturation: 0.716 versus 0.699, 0.747 versus 0.586, and 0.765 versus 0.628, respectively). The difference was significant for UA oxygen saturation (P <.05). All tests showed high sensitivity and negative predictive values, but low specificity and positive predictive values.
Conclusion: Low fetal oxygen saturation measured at birth seemed to be associated with low fetal pH and base excess values, but its predictive value for acidosis in an unselected population was limited, particularly if acidosis was metabolic.